Dynamics of Transport and Energy Systems. History of Development and a Scenario for the Future

A. Griibler, N. Nakicenovic, and A. Schafer

International Institute for Applied Systems Analysis Laxenburg, Austria

RR-93-19 December 1993

Reprinted from "Energy and Life" World Energy Council, 15th Congress, (20-25 September 1992, Madrid).

INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS Laxenburg, Austria

of the Institute, its National Member Organizations, or other organizations supporting the work.

Reprinted from "Energy and Life" World Energy Council, 15th Congress, 20-25 September 1992, Madrid.

Copyright @1992 World Energy Council.

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage or retrieval system, without permission in writing from the copyright holder.

Printed by Novogra.phic, Vienna, Austria..

Scenarios of future energy developments and of their likely environmental impacts, particularly in the transport sector, face a number of challenging questions. For example, as societies develop will there be a global homog- enization of lifestyles leading to higher demand for individual mobility and energy-intensive model activities as we see today in the industrialized coun- tries? Are there alternatives that would challenge this conventional wisdom that envisages rapid growth in energy consumption and emissions? And fi- nally, could we assess the importance of currently used technologies in t.t"te future, on the one hand, and the ultimate need for fundamentally new solu- tions, on the other?

This report analyzes the historical development of transport systems from canals to railroads, from horse-driven carriages to automobiles, and more recently to air transportation and new forms of high-speed ground transportation. The historically observed heterogeneity in the timing, du- ration, and adoption intensity of the diffusion of individual transport tech- nologies is then used to sketch a scenario of energy demand, and resulting carbon emissions from the passenger transport sector.

In this scenario, automobile and air transport in the industrialized coun- tries approaches saturation in the next two to three decades. For developing countries, however, the scenario suggests that despite their high growth po- tential, they may not necessarily mimic the high adoption intensity of indi- vidual transport modes prevailing in the industrialized countries. Combined with a continuation of historical rates of efficiency improvements, trans- portation energy demand and resulting carbon emissions could level off after the year 2000. Carbon emissions could be reduced significantly thereafter through the penetration of electric- and hydrogen-powered vehicles in the in- dustrialized countries. Such a scenario illustrates a possible future pathway

III

contrasting conventional wisdom and giving some indication of the direc- tion and time scale of technology and transport policy actions required to reconcile growing mobility with global environmental concerns.

Peter E. de Jdnosi Director

Dinamica de sistemas de trans,porte y de energia:

historia del desarrollo y escenario para el futuro GRUBLER, A., NAKICENOVIC, N., and SCHAFER, A.

International Institute for Applied Systems Analysis, Austria

1. Introduction Introduction Introducci6n

The history of transport systems is one of revolutions. Technological changes from the mail-coach to the airplane have transformed and extended the spatio-temporal range of human interaction patterns, leading to unpre- cedented levels of performance in terms of speed, quality of service, spatial division of activities, and integration of economic spaces. Similar dramatic changes in the energy system have accompanied (in fact made possible) the changes in transport technologies. In this report we provide a quantitative history of the transport system and draw some conclusions concerning possi- ble future developments and implications for energy use. Evidence of long- term regularities in the evolution, diffusion, and replacement of several fami- lies of technologies that constitute our transport system emerges, thus facili- tating a prospective look into the future.

The evolutionary envelopes analyzed reveal a process of gradual replace- ment of old by new systems along structured development trajectories that can be formalized mathematically by simple, biological growth and inter- species interaction models. Older systems are made obsolete through tech- nological advance and economic development, and new on~s are introduced that are better adapted to continuously changing social, economic, and

Submitted to "Energy and Life" World Energy Council Congress,20-25 Sept.1992, Madrid.

environmental boundary conditions. Above all it is the advancement of tech- nology that has denied or forestalled the original Malthusian resource deple- tion myth. Mankind has been able to modify and increase the size of its niche (including its spatial range) and to sustain a growing population at higher levels of economic well-being and mobility.

New transport technologies have been vital for economic development since the onset of the Industrial Revolution. In fact they have been so vital that economic historians have termed whole periods of economic development after various transport infrastructures, e.g., the "age of canals" in the first half of the 19th century, or the "railway ttnd coal era", the expansion of which ended with the Great Depression in the 1930s. The oil and automo- bile alliance was the symbol, and one of the main contributing factors, to an expansion period unprecedented in the economic history of mankind, but this ultimately will also come to an end. The turbulence and volatility wit- nessed since the 1970s may be an indicator of a deeper structural transfor- mation in the economy as a whole, and of the transport and energy sectors in particular.

Thus, the growth of individual transport systems is not a continuous process but time dependent. Based on historical analysis, we conclude that there was a time to build canals and railways, as there was a time to build high- ways and have increasing car ownership rates. However, despite the fact that the development of individual transport systems may be extremely suc- cessful over periods of several decades, any boom period will ultimately be followed by a structural discontinuity, a season of saturations. The expan- sion of a particular technological system reaches limits in terms of market saturation, social acceptability, and environmental constraints.

We suggest, again based on historical analogy, that it may be more creative to think about the opportunities generated by the transition to a new tech- nological regime, rather than to develop scenarios assuming further develop- ment along the trajectories and intensity levels of the previous, by now saturating and ultimately vanishing, modes of economic and infrastructural development. Based on this working hypothesis we sketch out a scenario for the future evolution of the transport sector and illustrate the implications of this scenario for energy demand, in particular for oil. The scenario postu- lates a forthcoming saturation of automobile diffusion in industrialized countries with lower adoption levels in the developing countries. At the same time, we consider the possibility of further growth in passenger tran- sportation, as well as the emergence of new transport modes, such as high- productivity super- or hypersonic aircraft fueled by methane or hydrogen, maglevs (magnetic levitation trains) based on superconductivity, or electric and hydrogen powered automobiles.

102

_ _

101.!- /f/Buses + .!-10,000

Figure 1. Growth of Mobility (Distance Traveled per Day per Capita), France (Griibler, 1990). ^.

Croissance de la mobilite individuelle en France.

Crecimeinto de movilidad en Francia.

Such a scenario may appear unrealistic with current technologies and insti- tutions, and in any case infeasible with a further intensification of tradi- tional transport modes and no new breakthroughs in energy efficiency and reduction of emissions. It is however consistent with the historical trend towards greater spatial range and increasing mobility in the world. Figure 1 shows that the increase in mobility, expressed as the average daily travel range, has been a pervasive historical development. Mobility increased through four orders of magnitude (by a factor of ten thousand): For exam- ple, the spatial range covered by the population of France appears to grow along a secular trend that increased from about 20 meters per day per cap- ita in 1800 to more than 35 kilometers today. These increases in mobility were primarily achieved by the introduction and growth of new transport modes, railways in the 19th century and buses and cars in the 20th century.

Although the increase in mobility may be sustained over several decades by growth in the output of the dominant transport system, as illustrated by the growth in railway traffic between 1860 and 1920 and by the growth of road

100d- -, I.,vr

I

10.1..). ""

f

(, 100

.I' ! "

-

Waterways

/

.' _j I

I

10.2 ^I 10

1800 1825 1850 1875 1900 1925 1950 1975 2000

traffic since WW II, there. are obvious limits to the further expansion of traditional transport systems. Current transport modes will ultimately have to be replaced by more advanced (in terms of productivity and environmen- tal compatibility) systems.

2. Growth of Transport Infrastructures

Croissance des infrastructures de transport

Crecimiento de la infraestructura de meaios de transporte

The first railways were constructed in the 1830s and extended the range, speed, and productivity offered by previous transport infrastructures, in par- ticular turnpikes and the elaborate system of canals constructed between the end of the 18th and beginning of the 19th century in a number of coun- tries. In time, elaborate networks of railway systems were built in North America and Europe. Together with railways, a new era of coal, steam, steel, and telegraph began. The great railway era lasted until the 1930s.

Figu.re 2 shows the diffusion of railroads in Europe, the United States, and in the "rest of the world" illustrating the extent to which their development was synchronized internationally. While the processes differ both in rates of growth and duration, there is a high degree of congruence in their ultimate saturation in the industrialized countries towards the 1930s.

The diffusion cluster of railway networks in Figu.re 2 that reached saturation by the 1930s accounts for almost 70 percent of all railroad lines constructed to date. This implies that the particular development trajectory (the "rail- way bandwagon") was not repeated by latecomers, i.e., by those countries that did not participate in the expansion pulse of the 19th century. In fact, the global railway network reached a length of about 1.3 million kilometers by the 1930s and has remained at that length ever since. Decommissioning of lines in the core regions (Europe and North America) was compensated for by new construction in the developing countries. In the core countries the railway bandwagon focused as it evolved, converging toward the satura- tion period. In other words, although the introduction of railroads occurred with great lags between the early and late adopters, such as between the UK and Scandinavia, the late adopters achieved faster diffusion rates with a pronounced catch-up effect converging towards saturation. This focusing of the expansion pulse of the railway era is most noteworthy when one consid- ers that the whole development process lasted about 100 years in the leading countries and only a few decades in the late adopters. This phenomenon is not unique to railways. The development of canal networks (in the UK, France and the US) saturated between the 1850s and 1860s, with the fol- lower countries catching up to the British "canal mania" initiated at the end of the 18th century.

300

-

USA~ '}...

1'-' .0' -..,

, .

--;::

, ..

, .

,. ,:

,,' ....

, .

, ..

" ..., ,,' ..."

.../ Rest of World

I": ~

I' ,:

, .

, ..

I' .0'

, .'

, .

, ..

,. ,0'

o .

,.' ....

, .

.,' ...

.'- ::.::: ~::.::.:~~: ' ...

1000 km 270 480 460 1210 250 410 640 1300

500

200

1925: Europe:

USA:

Others:

World:

1985: Europe:

USA:

Others:

World:

400

300

200 100

100

1800 1825 1850 1875 1900 1925 1950

Figure 2. Growth of Railway Networks, in 1000 km.

Croissance des reseaux de chemin fer.

Crecimiento de redes ferroviarias.

The absolute level of adoption is however much lower for the late adopters compared to the early adopters. The ultimately achieved railroad density is in general higher the earlier the railroads are introduced; leaders achieve the highest diffusion levels. Figure 9 shows this phenomenon. It demonstrates that the railway densities can be grouped within a declining "density envelope" as a function of the introduction date. This result shows that whereas the diffusion rate accelerates (i.e., the "catch-up" effect) for late adopters, the intensity of final adoption levels decreases. The intensity falls even further for those countries that introduced railways after the comple- tion of the expansion pulse in the core countries. This identifies an "oppor- tunity window" for the diffusion of pervasive systems like railways, which were the dominant transport system during the second half of the last cen- tury.

The advent of the automobile around the turn of the century initiated a new era in the development of transportation. Increased mobility became the symbol of industrial development along with oil, petrochemicals, electricity, telephone, and (Fordist) manufacturing. The flexibility offered by an

Railway density

km/100 km2

(1929)

year of maximum network size

~K:~92(8:~'~~::"{lg;:3)

US~r(~929)

\,_''''~~~.n

⁽¹⁹³⁸⁾

Europ.Russia (19Ij). · .A~!lPna (1955)

"~~, · Me)c1~\\(1975)

Canada (1974)-"t!__~.__ _ -";.<

USSR ~:::::;.':}f:;I:AM~tr~!t~,-"...

(1985) Brazil r19jgr~"~;'~a (1985) (1960)

10.0

1

0.1

1800 1820 1840 1860 1880

Network construction date (year network > 1% of final maximum)

1900

Figure $. Railway Network Density as a Function of Introduction Date, World (Grubler and Nakicenovic, 1991).

Densite des reseaux de chemin de fer en fonction de leur date d'introduction.

Densidad de redes ferroviarias dependiendo de la fecha de introducci6n.

individual mode of transport became affordable for a wider social strata, and only recently have the disadvantages of the automobile become socially transparent although they have been known for a long time. This percep- tion lag illustrates the extent to which the automobile became accepted as one of the preconditions for modern industrial development.

The growth of the automobile to the present number of about 420 million passenger cars registered worldwide went through two distinct diffusion phases marked by different growth rates divided by a structural discon- tinuity in the 1930s. The first phase was initiated during the last century when horse-driven vehicles were the predominant form of road travel.

Figure .4 illustrates the first, rapid phase of automobile diffusion as it replaced horse-drawn vehicles in the United Kingdom, together with a simi- lar substitution process in the railway transport system a few decades later

80 Horse carriages

&. hackneys

Steam locomotives

60

40

20

o

1900

4.

by Diesel and Electric Locomotives, U.K., in Share in Total Number of Rolling Stock.

Remplacement des vehicules equestres par l'automobile et des locomotives a vapeur par celles a diesel on electrique, Royaume- Uni.

Sustitucion de vehiculos (caballos por automoviles) y trenes (vapor por diesel y electricidad), Inglaterra.

when diesel and electric locks replaced steam propulsion in the rolling stock. 1 This example is typical of similar replacement processes throughout the world including the former centrally planned economies and developing countries. Technological change and substitution in road and railway vehi- cle fleets lasted in general between two to three decades.2

1 The fractional shares, F, are plotted as the share taken by the new competitor divided by the volume of the whole market or "niche", Le., the share of the cars in the number of all cars and horse-drawn vehicles together. This form of presentation reveals S-shaped substitution paths of old by new technologies and systems as formulated by Fisher and Pry (1971). In most cases, this process can be described by logistic functions.

2 The duration of the diffusion process is measured as the time that elapses between the achievement of 1 percent and 50 percent (and from 50 to 99 percent) of the complete replace- ment of old by the new or the acliievement of so-called market saturation. We call this meas- ure D.t. D.t also approximates the time required to grow from 10 to 90 percent of the saturation level.

density cars/l 000

USA~..

Diffusion rate lH years

1000 500

100 50

100 50

USAlolal ..

CANNZUSAPCYV MEX

~ ^..~

AUSGB

I

D P

~

SF AICS

J" ..

World WAN

10

10

-( _{I I I I I I I I I I I I I I}

1880 1920 1960 2000

5. Diffusion Rates and Ultimate Ownership Densities of Passenger Cars as a Function of Introduction Date, World (Griibler, 1990).

Vitesse de diffusion et niveaux d'adoption ulterieure de l'automobile en fonction de sa date d'introduction.

Tasa de difusion y densidad de propiedad de automoviles depen- diendo de la fecha de introduccion.

Figure

The adoption level of the automobile in different countries follows the same pattern as the observed saturation levels of railroads. Of the industrialized countries, the diffusion rate was slowest in the United States and Canada, faster in most of the European countries, and fastest in Japan. The adop- tion of the .automobile was even faster in developing countries, albeit that it started only recently and has led, at least for the time being, to relatively low levels of car ownership.

Figure 5 reports these results by showing the diffusion rate (6.t, given on the lower curve and labeled on the left vertical axis) as a function of the begin- ningof .the innovation diffusion on the horizontal axis. The band at the top of the figure gives the estimated automobile adoption levels (plotted on the right vertical axis in number of cars per 1,000 capita with the associated sta- tistical uncertainty bands), again as the function of the automobile intro- duction date. Thus, a pronounced acceleration of the diffusion rates can be

This result is consistent with spatial patterns of innovation diffusion. Ori- ginating from innovation centers that reach the highest adoption levels, the innovation generates additional gravity centers in space; however, the adop- tion levels remain lower in the peripheral regions compared with innovation centers. Early innovators achieve higher adoption levels and have the long- est diffusion or "learning" times. Peripheral regions catch-up so that the diffusion process focuses toward the saturation period, but the adoption level in the periphery remains much lower than in the leading centers.

Therefore whereas it is likely that road transport will continue to increase in the developing countries, any massive diffusion of car ownership, especially to levels anyway close to those of industrialized market economies, cannot happen "overnight". It is more likely to take several decades and extend well beyond the time horizon of the illustrative energy scenarios adopted here. In the meantime, additional growth in mobility outside the OECD region will have to rely to a large extent

-

-

3. Technological Substitution Substitution technologique Sustitucion tecnologica

The evolution of transport infrastructures can also be seen as a substitution process. Infrastructures can be viewed as systems that replace each other in time. Figure 6 shows the growth of transport infrastructure length in the United States as a substitution process3 and for comparison the equivalent substitution process in the Soviet Union.

The pattern of temporal changes is marked by a high degree of regularity and a quest for higher speed and productivity. The phase transitions in the infrastructure substitution in the former USSR, however, are lagged by a few decades when compared with the United States. For example, the domi- nance of railways lasted until the 1940s while in the United States it ended two decades earlier. The decline of canals occurred much later as well, while the growth of national airways follows the same path as in the United States. During the last decades, development of transport infrastructures in 3 The fractional shares, F, are plotted as the linear transformation,Le., FI(1-F), as the ratio of the market share taken by one infrastructureover the sum of the market shares of all other competing systems. Every competitor under/loes three distinct substitution phases: growth

h

turation, and decline as shown by the substitution path of railways (and later roads) w ich curve through a maximum from increasing to declining market shares (see Figure 6). A detailed description of the model is given in Nakicenovic, 1979.

Fj(l

-

10.2 1800

Fraction (F) 0.99

0.90

0.50

0.10

1850 1900

0.01 1950 2000

Figure 6. Substitution of Transport Infrastructures in the USA and the Former USSR, Fractional Shares in Length (Griibler and Nakicenovic, 1991).

Substitution des infrastructures de transport aux Etas- Unis et en ancien Union Sovietique.

Sustitucion de infraestructuras de transporte en EEUU y la an- tigua URSS.

the two countries has been converging. The rate of relative growth in the importance of road infrastructure and their saturation also appear syn- chronized. However, road infrastructures are used differently in the two countries: by individual road vehicles (cars) in the US and by collective ones (buses) in Russia. Thus, there is an increasing congruence and similarity in the structural and functional evolution of the transport system in the two countries. To a large extent this is also due to the fact that both countries have relatively low population densities and vast territories that modern transport systems must bridge in a matter of hours.

Figure 7 shows the actual usage of transport infrastructures. It gives the long-distance passenger modal split in the United States (top of Figure 7), the former Soviet Union and China (bottom of Figure 7). These examples are representative of the developments in most countries and illustrate a

passenger transport development portrayed a phase-shift. Rail and bus are virtually extinct in the United States, while in the former Soviet Union and China they are still an important means of transport. Thus, while many characteristics of the former USSR's intercity modal split changes lag by several decades compared to the US (with the exception of air travel growth), China's lag of some 20 years behind the USSR (note the different time axes on the bottom of Figure 7) is even l-arger. ^.

The above indicates that by the end of the c~ntury airways may become the dominant form of intercity travel in both the United States and the former Soviet Union. In both countries road transport currently has the largest share, albeit by automobile in one case and by bus in the other. The developments in the former USSR may serve as a guide for the modal split changes in China over the next decades. What is important is that the aver- age choice of different modes of passenger travel changes consistently and favors faster and more productive systems.

The slower modes recede to service fewer people over shorter distances (fewer passenger-kilometers) and low-value segments of transport. Their shares in passenger transport decline, resulting in the rationalization of the respective infrastructures. The least productive links are decommissioned, and the infrastructure declines. The diffusion of more productive and faster means of transport leads to increased range and connects ever larger areas into single functional entities. During the last two centuries villages merged into towns, towns into cities, and cities into metropolitan areas. Large urban corridors have evolved throughout the world, some approaching a hundred million inhabitants requiring high productivity transport systems.

4. Transport and Energy Systems Systemes de transport et de l'energie Sistemas de transporte y de energia

The substitution of older transport infrastructures for new ones and the high degree of synchronization in the diffusion of transport systems are not coincidental. They reflect strong linkages and similar pervasive changes in the overall evolution of the social and economic system denoting major techno-economic paradigm shifts. The diffusion of railways is linked with the spread of steel, coal and many other related technologies throughout the world. The same can be said about cars, the internal combustion engine, oil, petrochemicals, etc. Figure 8 shows the substitution pattern in primary energy sources in the United States. Similar paths of technological change

FRACTION F

let ,.,-8Itn1

0.99

0.90

f

-- 0.50

_~---~.~1 0.30

~---,.---_..--

AIR 10.10

BUS

---...---- 0.01 2000

1970 1980 1990

China

1970 1990 2010 Fraction F

0.99

10.1 +- Boat

I

I

10-2 1920

0.90

0.50

0.10

0.01

1940 1960 1980 2000

-

Figure 7. Substitution of Intercity Passenger Transport Modes in the USA (top) and the Former USSR and China (bottom, with shifted time axes), Fractional Shares in Inter-city Pass-km.

Substitution des modes de transport inter-urbain aux Etas-Unis, en ancien USSR et en Chine.

Sustituci6n de modos de trans porte inter-urbano para pasajeros en EEUU y la antigua URSS y China.

F/(1-F) 102

10' rR- 100

10'1

10.2 1950

F/(1

-

market share of each consecutive primary energy source are separated by about five decades. The feed requirements for horses and working-animals saturate in the 1870s, coal around 1920, and oil around the time of the

"energy crisis" of the early 1970s.4

Technological changes in transport and energy systems have been instru, mental not only for raising productivity and providing new services but also for improving energy efficiency. The primary energy substitution of Figure 8 can also be interpreted as a substitution process of energy end-use and

F/(1- F) 102

Fraction (F) 0.99

0.90

0.50

0.10

10-2

1800 1850 1900 1950

, 0.01 2000 .Oil and gas technologies. respectively.

Figure 8. Primary Energy Substitution, USA (Nakicenovic, 1987).

Substitution des sources d'energie primaire, Etat- Unis.

Sustituci6n de energia primaria, EEUU.

4 "Oil technology" in Fipure 8 represents all primary energy consumption that is associated with oil production including associated natural gas. This is equivalent to the inclusion of town gas in tlie consumption of coal. "Gas technologies" exclude production of natural gas from oil wells.

conversion technologies, from the sailing to the steam ship, from stationary coal-fired steam engines to electric motors, from horses to internal combus- tion engine powered cars. The effect on improved energy efficiency of this technological change has been dramatic. A unit of value added in the US is produced today with only 20 percent of the primary energy requirements some 200 years ago (Nakicenovic et at., 1989). This corresponds to a long- term improvement rate in energy intensity of about one percent per year.

5. Scenario of Car Diffusion

Scenario de diffusion de l'automobile Escenario de la difusi6n automovilistica

Given that the replacement of transport and energy technologies follows broadly similar patterns both in space and time, our scenario of global mobility development assumes a similar co-evolution of transport and energy systems in the future. A new generation of automobiles and aircraft is assumed to be associated with a new generation of vehicle fuels, namely electricity and hydrogen. The scenario also assumes that the absolute diffusion level of these new technologies would be lower in the lagging regions of the world than in those that lead the innovation process.

The scenario of the future growth in automobile diffusion is based on two main assumptions. First, it portrays a S-shaped diffusion pattern that leads to a saturation after the end of this century. This is estimated from the his- torical growth of the world automobile fleet up to the present number of about 420 million passenger cars. Second, the diffusion process is separated into two phases, each corresponding to different market niches and driving forces of the growth of the automobile (see Figure 4, Nakicenovic, 1986, and

Griibler and Nakicenovic, 1991). During the first phase, the automobile diffused rapidly into the market niches previously held by horse-drawn car- riages, where its comparative advantage was particularly high and led to average growth rates of about 30 percent per year. The second phase, which started after the 1930s and is still in progress, is characterized by substan- tially lower growth rates: The number of cars registered worldwide increased typically by five percent annually.

The two phases of automobile diffusion are shown in Figure 9. They are especially visible on the logarithmic scale (left side) with the inflection point occurring around 1930. The scenario extends to the year 2010 and is shown on the linear scale (right) of Figure 9. Since the actual diffusion of the auto- mobile in the world is still decades away from the estimated saturation level, the scenario is based on two alternative cases derived from the historical data

-

200 t*

100 .Uncertainty range

assuming logistic growth and 90% probability.

105

1900 1920 1940 1960 1980 2000 2020

Figure 9. Passenger Car Diffusion and Scenario, W orId.

Histoire et scenario futur de la diffusion de l'automobile.

Difusion y escenario de automoviles.

5.1. Cars intheYear 2010

L'automobile dans l'an 2010 Automoviles en el ano 2010

Table 1 summarizes the results by regrouping the 1985 values and scenario ranges (the best fit and high cases) for the year 2010 by broad geographical regions based on the selected sample of 21 countries. For a better com- parison with other projections we have also aggregated 12 of the OECD countries (representing 98 percent of all passenger cars registered in the OECD) as a proxy for more developed countries, and the residual to the world total as a proxy for the sum of less developed (LDC) and (former) centrally planned economies (CPE). Table 1 shows the high degree of heterogeneity in the ultimate adoption levels among different regions of the world. The differences are generally a function of the diffusion time. Figure 5 illustrated that this is indeed the case and the scenario demonstrates the possible consequences of such a development in the future. The early adop- tion leads to higher diffusion levels and thus to higher and more pervasive automobile ownership. The scenario suggests an increase in the global

I.

(linear Icale)

/-

registered.

(logarithmIc Icale)

2nd Phase

+⁴⁰⁰

'/

ht Phase

I

D

, -- I.

109 J,

World

OEC , ,

== OECD -

,

--'-

-

.-- ^.-

.!. 300 ... USA

USA . "'<...."

10' J

passenger car fleets of between 21 and 38 percent compared with 1985 lev- els. The growth is higher in the developing regions both in absolute and in relative terms, and in the high case the car fleet actually doubles until the year 2010 outside the OECD region to a level of about 150 million cars.

Table 1. Scenario of Passenger Car Registration for the Year 2010.

Scenario du nombre de voitures immatriculees en l'an 2010.

Escenario de matriculas de automoviles para el ano 2010.

This scenario is lower than other projections in the literature, especially for developing countries. Often, pervasive adoption of the automobile is thought to lead to similar ownership densities throughout the world. In contrast, the scenario does not rely on the motorization trends of the indus- trialized countries as a guide of what might occur in the developing regions.

Instead, the rationale for the relatively low levels of automobile adoption in the world is based on the possible saturation of car densities in urban areas, an increase in air transport for longer journeys, the availability of new high- speed ground transportation systems, and perhaps also the increasing environmental problems associated with the further expansion of automo- biles. The high urban population growth rates in developing countries calls for the construction of efficient mass transit systems for short-range travel.

For long-distance travel, the growth potential for air transportation appears to be more consistent with the likely future developments in these countries than a linear extrapolation of the transportation trends of industrialized countries.

1985

--

...Cars/

Region 106 cars 106 pop. 1000 106 carso 106 pop. 10000

North Americah 147.2 264.7 556 169.4-187.4 303.1 559-618

10 OECD Countries' 141.9 427.0 332 153.7-169.1 458.6 335-369

4 CPEd 11.7 80.5 145 16.5-21.5 85.7 193-251

5 NICs' 22.6 291.5 78 39.0-69.8 397.6 98-176

World 364.8 4838.5 75 439.5-501.6 70]5.9 63-71

12 OECD Countries 289.1 691.7 418 323.1-356.5 761.7 424-468

% of 1D0rld 79% ]4%

-

LDCs and CPEs 75.7 4146.8 18 116.4-145.1 6254.2 19-23

% of world 21% 86% - _{27-29% 89%}

8Ra.uge corresponds to best fit and high ca.serespectively.

.USA and Canada.

· Australia, Austria, France, FRG, Italy, Japan, Spain,Sweden,NewZealand,and the UK.

"Cuchoslovalcia,former GDR, Hungary,and Poland.

· Argentina, Brazil, Mexico, Taiwan, and Venezuela.

varying degree of infrastructural development, etc. Countries like Argen- tina, Brazil, Mexico or South Korea have similar GNP per capita (of about 2,000 1982 US$ per capita in the Latin American Countries to about 2,500 1982 US$ per capita in South Korea) whereas in 1986 their car densities ranged between between 64 and 124 cars/l,OOO inhabitants in the Latin American countries and 16 cars/1,000 in South Korea. According to the scenario, the diversity in automobile ownership would continue to be a characteristic feature of both developing and more developed countries.

...

5.2. Scenario of Air Transport Scenario du transport aerien Escenario del transporte aereo

The long-term evolutionary changes in transport systems indicate that fas- ter modes are favored in long-distance traffic. A logistic function fitted to the actual data suggests that the inflection point in the growth of air carrier operations worldwide occurred around 1980. Figure 10 shows the statistical uncertainty bands associated with this estimate. The saturation level is estimated between 240 and 340 billion ton-km (passenger plus freight traffic) with 90 percent probability. The upper value of this range appears more likely since growth has barely proceeded beyond the estimated inflection point. The air transport scenario also includes the best fit and the high values as the possible volume of future global air transportation.

Contrary to individual automobile transport, the market volume growth in air traffic was sustained by the significant productivity increases of aircraft rather than by increases in the commercial aircraft fleet. Furthermore, much of the traffic is allocated to the most productive aircraft operating among the large hub airports, while other aircraft constitute the feeder and distribution systems for destinations with a lower volume of traffic. Over the last five decades the productivity of aircraft has increased by about two orders of magnitude. For instance, the DC-3 was introduced in 1935 with a performance of about 7,400 passenger-km/hr (21 passengers at 350 km/hr) whereas the Boeing 747 (B-747), introduced in 1969, has a performance of about 500,000 passenger-km/hr (500 passengers at 1,000 km/hr). The larg- est Jumbo will probably carry almost 700 passengers. The B-747 family is therefore about a hundred times as p,roductive as the DC-3 was 50 years ago (Nakicenovic, 1988). The volume of all airline operations has continued to increase since the introduction of the B-747 while its productivity increased only marginally through "stretching". Most of the subsequent technological development focused on gradual improvements such as noise reduction and

109 t-km 400

320

o

1930 1940 1950 1960 1970 1980 1990 2000 2010 2020

---

" , , '"

'"

,, 240

160

eo

Figure 1O. Air Transport Scenario, W orId.

Scenario du transport aerien.

Escenario del trans porte aereo.

increased fuel efficiency. Unavoidably, this leads to an impression of dimin- ishing returns in the further advancement of airliners and, consequently, the possibility of an approaching performance saturation of the current genera- tion of aircraft technologies. If we consider current airport overcongestion and the inconvenieI).ces associated with an increase in the size of an aircraft, the likelihood of developing hyperjumbos with capacity for a few thousand passengers is slim.

Accordingly our scenario of future air transport is lower than most aircraft manufacturers' forecasts (e.g., Boeing, see Steiner, 1989) not only because it employs a non-linear model but also because it takes into account the limits for further productivity improvements to the existing generation of aircraft.

With aircraft productivity saturating, future growth in demand could only be achieved by an increase in the number of aircraft in operation. Again, this appears unlikely in view of the already overcongested airspaces and air- ports in most developed countries. A second possibility out of this impasse would be a quantum leap in technology, similar to the introduction of the jet engine.

decades. Cruise supersonic and possibly air-breathing hypersonic aircraft connected to high productivity ground transport systems such as maglevs might be the only technologies now on the horizon to provide the growth potential that could be required in the next expansion phase of worldwide passenger travel and freight services.

6. Implications on Energy Demand and Garbon Emissions Implications pour la demande d'energie et les emissions C02 Consecuencias de la demanda de energla y emisiones de carbono Saturation of passenger car and air transport would have repercussions for energy demand and associated emissions. The scenario envisages passenger car growth of 20 to 40 percent and air transport growth of 60 to 80 percent by the year 2010 compared with 1985 levels. Figure 11 illustrates the energy demand of this scenario. Estimates are based on the medians between the best fit and high cases, i.e., passenger car growth of about 30 percent and air transport of about 70 percent. Three alternatives are shown: one based on 1985 automobiles and aircraft efficiencies; the second on "evolutionary"

efficiency improvements of about one percent per year; and a third that additionally envisages technological changes in the fleet resulting from the diffusion of electric and hydrogen propelled vehicles.

The final energy demand would level off at about 0.9 billion toe (40 109 GJ) per year if.the 1985 vehicle efficiency remains unchanged until 2010 (Figure 11). About 80 percent of the fuel would be consumed by cars and the remainder by aircraft. Carbon emissions from passenger cars and aircraft would also increase to eventually level off at a value of around 750 million tons carbon. This appears to be quite unlikely however as fuel efficiencies are certainly going to improve. Historically, fuel efficiencies in the transport sector have improved at a rate of about one percent per year, although in some cases like aircraft they are two to three times higher.

A value of about one percent per year improvement is adopted in the "evo- lutionary" efficiency alternative. The resulting automotive final energy demand by 2010 would basically be the same as in 1985, whereas air tran- sport fuel demand would be some 30 percent higher (Figure 11). After 2010, fuel demand would decrease due to the projected saturation combined with a continuation of fuel efficiency improvement trends. Under such assump- tions, energy requirements (and carbon emissions) by 2030 would drop to the levels that prevailed in the late 1970s.

Figure 11. Final Energy Demand for Transport, and Scenario, World.

Scenario de la demande en energie finale des transport.

Demanda final de energia para transporte y escenario.

The third alternative investigated combines the "evolutionary" efficiency improvement trends with the introduction of advanced hydrogen and/or electricity powered ground vehicles and hydrogen propelled aircraft. The industrialized countries are foreseen to take the lead in the development and diffusion of such a new generation of transport vehicles which combine significantly reduced fuel requirements and zero-carbon energy vectors. The replacement of conventional vehicles by electric cars or by cars equipped with hydrogen fuel cells, and of conventional jet-fuel aircraft by hydrogen powered ones offers a large potential for reducing final energy demand and, even more, carbon emissions.5

5 We. only discuss final energy demand due to the uncertainties of the future fuel mix and efficiency of electricity and hydrogen production. Moreover, ~rimary energy requirements

J

calculated via substitution equivalents of fossil energy inputs) are a rather problematic in ica- tor,in the context of renewa5le (solar or biomass) and carbon-free (hydro and nuclear) energy optIOns.

- ⁵⁰ -

Past Scenario w

.., 0

cr.

to:)

Energy requirements 1.0

Q cr.

40 for passenger cars

---

-

^-

c 0.8 ..-

QJ C

E ₃₀ Evolutionary

QJ

QJ

0.7 E

..

.S ..

CT .S

QJ 20 0.5 C"

.. _GJ

0.4 ..

.. »

0.3 ^b.O

QJ ..

C QJ

QJ 10

0.21 C

i6 ^QJ

0.1

-

u: c ^c

0 0 u:

1940 1960 1980 2000 2020

Year

hydrogen powered passenger vehicles is also assumed to start within techno- logically advanced and affluent countries with high passenger car densities (Le., the OECD region) due inter alia to policies for limiting local (ozone), regional (acidic precipitation) and global (C02) environmental impacts from the transport sector. Based on the historical replacement dynamics of the vehicle fleets, this new generation of vel),icles is assumed to diffuse throughout the world within three decades. Due to the much higher end-use efficiency, final energy demand (by 2030) of this scenario is only a third compared with the first alternative (frozen E!-t1985 efficiency). Overall, this alternative leads to 2.5 percent per year reduction in energy requirements per vehicle kilometer. Carbon emissions are even lower with some 150 mil- lion tons of carbon and are eliminated altogether in the OECD region.

These illustrative cases indicate that scenarios based on diffusion theory and the dynamics of technological change result in future development of indivi- dual transport with lower energy requirements and emissions. Market saturation and increasing awareness of the negative social and environmen- tal externalities associated with further intensification of the dominant tech- nological regime start to block further adoption. At the same time, incre- mental technological change concentrates on improving the performance of existing technologies via efficiency improvements and environmental "add- on" technologies. As such, Figure 11 provides a clear illustration of the significant impacts on oil demand of scenarios combining a forthcoming, saturation in the diffusion of automobiles and air transport with continued efficiency improvements.

Ultimately, however, small and incremental improvements have to yield to revolutionary changes in transport systems in order to lower environmental impacts and to respond to the evolving demands for mobility of people and goods. Perhaps the current "information revolution" bears witness to the almost unsatiable demand for improved interaction between human beings.

It is likely that communication technologies may substitute for some passenger travel. However, there are many indications that improved com- munication, in turn, also translates into a demand for significant additional face-to-face communication and thus more travel. Fulfilling such potential demands in an environmentally compatible manner will perhaps be the big- gest technological challenge in the decades to come. In the meantime, efforts have to continue to make the current generation of transport techno- logies more efficient and productive.

BibHography BibHographie BibHografia

Borderon, D., 1989, The Potential for Increased Efficiency of Energy Use;

Masterminding Tomorrow's Technologies, Group Planning, Shell Inter- national Petroleum Co. Ltd., London, UK.

Fisher, J.e. and Pry, R.H., 1971, A simple substitution model of technologi- cal change, Technological Forescasting and Social Change 3:75-88.

Griibler, A., 1990, The Rise and Fall of, InfrastrtLctures: Dynamics of Evo{tL- tion and Technological Change in Transport, Physic a Verlag, Heidel- berg, FRG.

Griibler, A. and Nakicenovic, N., 1991, Evolution of Transport Systems: Past and Future, RR-91-8, International Institute for Applied Systems, Analysis, Laxenburg, Austria.

Marchetti, C., 1987, Infrastructures for movement, Technological Forecast- ing and Social Change 32(4):373-393.

Nakicenovic, N., 1979, Software Package for the Logistic Substitution Model, RR-79-12, International Institute for Applied Systems Analysis, Laxen- burg, Austria.

Nakicenovic, N., 1986, The automobile road to technological change, diffusion of the automobile as a process of technological, substitution, Technological Forecasting and Social Change 29:309-340.

Nakicenovic, N., 1988, Dynamics and replacement of U.S. transport infras- tructures, in: J.H. Ausubel and R. Herman, eds., Cities and their Vital Systems, Infrastructure Past, Present and Future, National Academy Press, Washington, DC, USA.

Nakicenovic, N., Bodda, L., Griibler, A., and P .V. Gilli, 1989, Technological Progress, Structural Change and Efficient Energy Use: Trends World- wide and in A ustria, International Institute for Applied Systems Analysis, Laxenburg, Austria.

Schafer, A., 1992, Carbon Emissions in the Passenger Transport Sector:

Technology and Alternative Fuels, WP-9204, International Institute for Applied Systems Analysis, Laxenburg, Austria.

Schumpeter, J.A., 1935, The analysis of economic change, Review of Economic Statistics 17:2-10.

Victor, D., 1990, Liquid Hydrogen Aircraft and the Greenhouse Effect, WP- 90-02, International Institute for Applied Systems Analysis, Laxenburg, Austria.

Zahavi, Y., 1981, The UMOT-urban Interactions, DOT-RSP A-DPB 10/7, U.S. Department of Transportation, Washington, DC, USA.

evolutionary envelope is analyzed as a process of the gradual replacement of old by new systems leading to improvements in performance and quality of services that increased range and speed, and connected ever expanding terri- tories into a single functional system. Great l~aps in the advancement of transport systems have allowed to cover greater" distances in a unit of time, from traditional villages and preindustrial cities connected in one day by walking, to almost the whole world being connected in one day by aircraft.

New technologies are not adopted instantaneously, their diffusion takes time, in many cases several decades. There is great diversity in the timing, duration and ultimate application density levels of technologies among countries, economic sectors or different social strata. These patterns can also be observed in the way that new technologies succeed old ones. In those countries that introduce new systems early, diffusion is often pervasive and characterized by a long, sustained period of development. In other countries where diffusion is either blocked or occurs later, it can be significantly faster, "catching-up" with the leading countries. However, as the development time is shorter, the ultimate diffusion levels are also lower.

Consequently, railroad networks are smaller in those countries where they were introduced later; similarly, automobile ownership is also lower com- pared with the leading countries. Thus, ultimate diffusion (saturation) lev- els are frequently lower in the follower countries.

The above forms the basis for a scenario of the further diffusion of current, and the introduction of new, transport systems. It explicitly considers heterogeneity in economic, social and spatial structures and in varying development stages. Consequently, the ultimate diffusion levels of current technologies in the scenario differ widely between countries as a result of their differences in initial conditions, degree of economic development, spa- tial structures, etc. For example, automobile ownership and car fleets are seen to be rather close to saturation in the industrialized countries. The growth potential in follower countries is high, yet ultimate car ownership levels will probably remain significantly below those achieved in the leading countries. Thereafter, the diffusion of a number of new technologies might lead to a renewed growth phase: Maglevs and hypersonic aircraft are possi- ble candidates. This, however, would represent a revolutionary change. In the meantime, incremental and cumulative innovations will make the future derivatives of current automobiles and aircraft much more efficient and more productive.

- -